The explosion of computing power and integration has led to a production of a continuous and massive quantity of electronic data. Such data can be reflective of computer communication or operations and/or for real-world happenings. Frequently, hidden amongst the data, are patterns that can be used to improve computer systems, business decisions, etc. The patterns can be more confidently determined and used when large amounts of data are available. However, collecting such data is itself an undertaking, especially in cases where data is to be collected from multiple complex sources. Each source can utilize different interface features, requiring data collection to be performed in a source-specific manner and making it difficult to predict how a data-collecting effort will progress. Thus, data collection is frequently surrounded by failed and prolonged attempts.